Abstract:

In this study, the use of Protective Glasses, which is one of the Occupational Health and Safety measures, was determined by Image Processing method. The glasses images taken through the placed camera were determined by Image Processing and Deep Learning. Throughout the study, Python Programming Language, Google Colab Platform, OpenCV library were used and object recognition algorithms were used. First of all, the obtained Goggles photos were collected in a file, and the labeling process was carried out on the MakeSence.Ai platform. Coordinate information was obtained by taking the detection documents of each photo, the results were divided into two as test and training data at a certain rate and transferred to the Google Colab Platform with the help of Darknet. The transferred data was trained with the YOLOv4 Algorithm, all results were run through the Python OpenCV library and their accuracy was determined. This study, which reached appropriate results by using Image Processing and Deep Learning methods, brings an innovative perspective in terms of running the YOLOv4 Algorithm on the Google Colab Platform and, as a result, facilitating the control of full-time Occupational Health and Safety measures in enterprises.

Keywords: